Aggregation refers to the assembly of proteins into nonphysiological higher order structures. While amyloid has been studied extensively, much less is known about amorphous aggregation, a process that interferes with protein expression and storage. Free arginine (Arg+) is a widely used aggregation inhibitor, but its mechanism remains elusive. Focusing on myoglobin (Mb), we recently applied atomistic molecular dynamics (MD) simulations for gaining detailed insights into amorphous aggregation (Ng J. Phys. Chem. B 2021, 125, 13099). Building on that approach, the current work for the first time demonstrates that MD simulations can directly elucidate aggregation inhibition mechanisms. Comparative simulations with and without Arg+ reproduced the experimental finding that Arg+ significantly decreased the Mb aggregation propensity. Our data reveal that, without Arg+, protein-protein encounter complexes readily form salt bridges and hydrophobic contacts, culminating in firmly linked dimeric aggregation nuclei. Arg+ promotes the dissociation of encounter complexes. These "unproductive" encounter complexes are favored because Arg+ binding to D- and E- lowers the tendency of these anionic residues to form interprotein salt bridges. Side chain blockage is mediated largely by the guanidinium group of Arg+, which binds carboxylates through H-bond-reinforced ionic contacts. Our MD data revealed Arg+ self-association into a dynamic quasi-infinite network, but we found no evidence that this self-association is important for protein aggregation inhibition. Instead, aggregation inhibition by Arg+ is similar to that mediated by free guanidinium ions. The computational strategy used here should be suitable for the rational design of aggregation inhibitors with enhanced potency.